Hacker News Reader

I suspect that the frequent

“I apologize, an an ethical AI, I cannot discuss…”

will ruin the illusion.

Using one of open source models without these restrictions would be much better.